GPT-5.1 represents a meaningful step forward in LLM capabilities. Three key improvements stand out:
1. Engine Segmentation & Personality Presets
The ability to segment different engine types with distinct personalities is genuinely useful. As a GTM builder, this means I can deploy contextually-optimized responses without extensive prompt engineering overhead.
2. Superior Instruction Following
The model now handles multi-step constraints simultaneously. Complex instructions that previously required 3-4 iterations now work on the first try. This directly reduces latency in production systems.
3. Improved Tone Adaptation
GPT-5.1 understands conversational context better. It shifts tone appropriately based on input, which matters more than people realize for enterprise adoption. Technical superiority loses to human-like interaction every time.
The Real Unlock: This isn't a revolutionary leap. It's a solid incremental advance that compounds when deployed at scale. The real advantage goes to teams building on top of this—not those claiming AGI is here.
Raycast
What's new in the Codex app
🔀 Built-in worktrees
Enable multiple agents to work without conflicts
• Use isolated worktrees in same repository
• Review clean diffs and provide inline feedback
📋 Plan mode
Type /plan to go back and forth with Codex
• Create thorough plans before you start coding
• Iterate on your approach with the agent
🗣️ Personalities
Choose the interaction style that fits your vibe
• Use the /personality command across all surfaces
• Pick a pragmatic or conversational style
🚀 Skills
Connect the tools you already use and go beyond writing code
• Deploy to @Vercel , fetch from @Figma , manage @Linear, and more
• Bundle your workflows into reusable skills
🔁 Automations
Delegate repetitive recurring tasks in the background
• Set up tasks for issue triage, failure reports, and more
• Combines skills and custom instructions to run on a schedule
Available on macOS, with Windows coming soon.
To celebrate, Plus, Pro, Business, Enterprise, and Edu users have doubled rate limits across the Codex app, CLI, IDE extension, and cloud.
Codex is also included in Free and Go plans for a limited time.
Congrats on the launch! 🚀 OpenAI remains a foundational, production‑grade platform for building and scaling serious AI products.
Zivy
Interesting stuff!
Cool to see OpenAI coming up with new alternatives - looking forward to testing it out
The worktrees feature is the sleeper hit here. Most developer workflows break down when you need context switching - you're mid-refactor on one branch, but a critical bug comes in. With isolated worktrees, you can spin up a separate agent instance without losing state on your current work.
The skills architecture is also smart. Rather than trying to make Codex do everything natively, connecting to Vercel/Figma/Linear means the agent can actually complete end-to-end workflows. Deploy, then open a PR, then update the Linear ticket - without leaving the context window.
Curious about failure handling though - when an automation runs overnight and hits an edge case, what's the recovery flow? Does it queue for human review or attempt self-correction first?
Good job! Has someone done a thoughtful comparison with Claude Code?